Congestion Pricing Based On Trip Decision And Uncertain Demand Forecasting

Traffic congestion has been a chief problem in many large cities in the world today. Improving traffic supply is the traditional method for solving the problem of urban traffic congestion. It is declared in the Adams law that the general tendency of traffic demand outweighs traffic supply and it is difficult to handle the congestion problem only relying on improving traffic supply. Combining with advantages of traffic demand management and traffic management systems, congestion pricing is the main economic means to solve the traffic problem, which can improve transportation efficiency and efficiently utilize traffic sources.Firstly, the theory of marginal cost under congestion is introduced and quantitative description of congestion pricing as well as its limitations is pointed out. One of the main limitations is the neglect of individual trip decision behaviors so that the toll strategy cannot meet the need of reality. Considering trip decision behaviors, models for UE and SUE are given in this thesis.Secondly, congestion pricing model is studied on the basis of traffic network equilibrium theory.The author presents the generalized cost based on an overall consideration of time cost, travel cost and congestion tolls. Then combining individual trip decision, a congestion pricing model with generalized cost is built and the directly search algorithm is proposed based on step accelerating method and penalty function method. Finally, taking uncertain demand forecasting into further consideration, the congestion pricing model is studied by sensitivity analysis in conditions of long uncertain demand forecasting. Thus, a congestion pricing model with uncertain demand forecasting is given, which is solved using the heuristic algorithm based on sensitivity analysis.